Skip to content

Latest commit

 

History

History
5 lines (5 loc) · 445 Bytes

README.md

File metadata and controls

5 lines (5 loc) · 445 Bytes

Churn Prediction

This is a practice on Kaggle Telco Customer Churn dataset (https://www.kaggle.com/blastchar/telco-customer-churn).
The objective is to predict whether a customer will churn using the variable.
The target class is imbalanced, therefore I Used f1_score & recall rate for model tuning and evaluation.
I started with a simple EDA on numeric/categorical variables and applied 6 different machine learning models.